ATTACHMENT SECURITY AND INTERNET GAMING DISORDER 2
The Effect of Stress Susceptibility on the Relationship Between Attachment and Internet Gaming
A Plan B Research Project
SUBMITTED TO THE FACULTY OF
UNIVERSITY OF MINNESOTA
BY
Imane Ait Daoud
IN PARTIAL FULFILLMENT OF THE REQUIREMENTS
FOR THE DEGREE OF
MASTER OF ARTS
Chair Dr. Rebecca Gilbertson, Ph.D.,
Dr. Molly A. Harney, Ph.D
Dr. Kathy Dowell, Ph.D.
April, 2020
ATTACHMENT SECURITY AND INTERNET GAMING DISORDER 3
© Imane Ait Daoud, 2020 ATTACHMENT SECURITY AND INTERNET GAMING DISORDER 4
Acknowledgements
Firstly, I would like to thank the University of Minnesota for providing me with one of
the best possible levels of education in the country for both my undergraduate and graduate
degrees. The University has provided me with resources such as access to the best faculty,
opportunities in research as well as personal and professional growth opportunities which have
shaped me as an individual, student and professional.
I would also like to thank my family and my friends for supporting me throughout my
journey to obtain a higher level of education and for sacrificing on my behalf to ensure the
manifestation of my dream. To my parents and husband, I am especially grateful for your
unconditional love and support.
To the members of my plan b committee Dr. Molly A. Harney and Dr. Kathy Dowell, thank you for your support, insight, and feedback which made this research project possible. The
amount of time and effort you have given this project did not go unnoticed.
To Dr. Gilbertson, thank you for your support and for believing in me even when I found
it hard to believe in myself. Thank you for all the lessons you’ve taught me academically, professionally, and personally through your mentorship. My experience as your student in your lab has been invaluable and I will take everything I’ve learned from you and carry it with me for
the rest of my journey through school, work and beyond.
ATTACHMENT SECURITY AND INTERNET GAMING DISORDER 5
Abstract
The prevalence of gaming in the United States has increased dramatically in the last decade (Entertainment Software Association [ESA], 2019). Excessive gaming, however, has also become problematic, with negative effects on relationship development, emotional wellbeing, stress susceptibility and more (Kaess et al., 2017; Monacis et al., 2017; Griffiths, 2005). Excessive gaming or the pathological use of internet games is referred to as Internet Gaming Disorder (IGD) (American Psychological Association [APA], 2013). While there is an abundance of research assessing the consequences of excessive gaming, there is still much left to discover about the etiology and development of Internet Gaming Disorder (IGD). Recently, the connection between attachment security and IGD has been increasingly explored, with several studies suggesting a positive correlation between insecure attachment and IGD (Eichenberg et al., 2017; Tavakoli et al., 2014; Benarous et al.,2019). This study sought to expand on the current literature by assessing the relationship between both attachment and stress susceptibility, and IGD and stress susceptibility. Additionally, a potential mediation model was also assessed. Participants (N = 423) completed self-report assessments regarding gaming addiction, attachment, perceived stress, trauma, and other demographics. Results indicated that significant differences between attachment groups on Internet gaming disorder scores (F (3,419) = 70.80, p < .001, w = .33), with the secure group (M=6.13, SD = 4.75) reporting the fewest IGD symptoms. Secure individuals (M=51.21, SD = 23.90) also reported significantly lower levels of perceived stress than insecure individuals, F (2,418) = 61.61, p < .001. A Tukey’s post-hoc analysis showed that the secure group differed significantly from the preoccupied group (p <.001), and the Anxious-Avoidant group (p <.001) but not the Dismissive group (p = ns.). IGDS scores were significantly associated with Attachment Anxiety (r = .63, p < .01), Attachment Avoidance (r = .26, p < .01), and ACE trauma scores (r = .69, p <.01). Lastly, a mediation analysis showed that while controlling for perceived stress, attachment type was still a significant predictor of gaming addiction scores, b = -3.17, SE = .72, consistent with partial mediation. The findings of this study indicate that attachment security, as well as stress, may have significant implications in the understanding of the development and etiology of gaming addiction. Further research is called for on the pathways through which attachment may influence gaming addiction and treatment of psychopathology. Keywords: internet gaming disorder, attachment, insecure attachment, secure attachment, stress susceptibility ATTACHMENT SECURITY AND INTERNET GAMING DISORDER 6
Table of Contents
List of Tables ...... 7
Introduction ...... 8
Internet Gaming Disorder ...... 8
Attachment Theory ...... 11
Attachment and IGD ...... 13
Attachment and Susceptibility to Stress ...... 15
Stress Susceptibility in Insecure Adults ...... 16
IGD and Susceptibility to Stress ...... 18
The Current Study ...... 19
Hypotheses and Predictions ...... 19
Methods ...... 20
Study...... 20
Population ...... 20
Primary Measures ...... 22
Procedure ...... 25
Statistical Analysis ...... 26
Results ...... 27
Discussion...... 34
Conclusion ...... 38
References ...... 40
ATTACHMENT SECURITY AND INTERNET GAMING DISORDER 7
List of Tables Descriptive statistics for gaming, stress, and demographic measures by Table 1. attachment. 48
Table 2. Zero-Order Correlations 49
Hierarchical regression Attachment Table 3. and IGDS 50
Hierarchical regression Attachment Table 4. 51 and Perceived Stress
Hierarchical regression Perceived Table 5. Stress and IGDS 52
Hierarchical regression Perceived Table 6. Stress, Attachment and IGDS 53
PROCESS Mediation Results 54 Table 7.
Table 8. Mediation Analysis Results 55
ATTACHMENT SECURITY AND INTERNET GAMING DISORDER 8
The Effect of Stress Susceptibility on the Relationship Between Attachment and Internet
Gaming Disorder Symptoms
Gaming, specifically internet gaming, is one of the fastest growing industries in technology today (Entertainment Software Association [ESA], 2019). Games are present in many households with easy access available to children, adolescents and adults throughout North
America. According to the ESA, in the last year alone, the gaming industry made about 43.4 billion dollars in profits (ESA, 2019). According to a recent survey by the ESA, approximately
164 million American adults play video games with the average age of gamers being 33. This same survey found that about 65 % of Americans report gaming regularly, with 75% of households reporting having at least 1 gamer (minimum of 3 hours a week) (ESA, 2019). This number has increased by 20% since 2015 (Forsans, 2015). Of the individuals surveyed, over half reported using video gaming as a method of connecting and making friends (Forsans, 2015).
Although the gaming industry often cites many positive effects of gaming such as “stress relief”,
“relaxation” and “mental stimulation”, many studies have also found that gaming behaviors, when excessive, can have severely detrimental effects on individuals (ESA, 2019; Parzer, Mehl,
Weil, Strittmatter, Koenig, 2017; Sugaya, Shirasaka, Takahashi & Kanda, 2019; Monacis, Palo,
Griffiths & Sinatra, 2017; King & Delfabbro, 2016). While there is an abundance of research assessing the consequences of excessive gaming, there is still much left to discover about the etiology and development of Internet Gaming Disorder (IGD).
Internet Gaming Disorder
Excessive gaming or the pathological use of internet games is referred to as Internet
Gaming Disorder (IGD) (American Psychological Association [APA], 2013). IGD is a ATTACHMENT SECURITY AND INTERNET GAMING DISORDER 9 behavioral addiction in which the persistent and recurrent use of internet games leads to a significant level of impairment and distress (APA, 2013). IGD has recently gained greater attention in the clinical realm of psychology. Though there is yet to be an official diagnosis for internet addiction, IGD has recently been included as a tentative disorder in the Diagnostic and
Statistical Manual of Mental Disorders (DSM-5) (Monacis et al., 2017; Kaess et al., 2017).
However, currently, gaming disorder is included in The International Classification of Disease,
Eleventh Revision (ICD-11) (World Health Organization [WHO], 2018). The criteria specified in the DSM-5 helps create consistency in criteria used in studies pertaining to IGD (Sugaya et al.,
2019).
IGD Criteria.
Griffiths’ model of addiction is comprised of six criteria including salience, mood modification, tolerance, withdrawal, conflict and relapse (Griffiths, 2005). According to Griffiths
(2005) salience refers to the process through which a behavior takes over a person’s cognitions, emotions and behaviors (Griffiths, 2005). Mood modification is the process through which a behavior or substance influences an immediate effect towards one's mood state. Tolerance is the increased need overtime to engage in a behavior in order to achieve the desired effects; withdrawal refers to negative symptoms experienced as a result of addictive behaviors (Griffiths,
2005). Lastly, conflict refers to the cognitive dissonance experienced when there is a conflict of interest between the addictive behaviors and relationships and responsibilities (Griffiths, 2005).
Increases in time investment, emotions of anxiety, depression, and denial similar to those seen in internet addiction, gambling addictions and substance abuse disorders have been noted in IGD
(Monacis et al., 2017). These six criteria can be measured using the Internet Gaming Disorder ATTACHMENT SECURITY AND INTERNET GAMING DISORDER 10
Scale (IGDS; Lemmens, Valkenburge & Gentile, 2015), which is consistent with the DSM-5
IGD criteria as well as Griffiths’ model of addiction.
Development of Disorder
Studies have also identified several individual factors that may contribute to the development of problematic gaming, including psychological and social problems, impulsivity
ADHD, sensation-seeking tendencies, and cognitive and family factors (Sugaya et al., 2019).
IGD in adolescents is often found to be comorbid with several other disorders including depression, ADHD, conduct disorder and anxiety disorder (Sugaya et al., 2019). Affected individuals tend to exhibit higher levels of self-harm behaviors and report a greater amount of interpersonal difficulty. In comparison to healthy individuals, adolescents and adults with IGD also show higher degrees of impulsivity. In addition, in males with IGD, sensation seeking, defined as the seeking of complex, and intense sensations regardless of physical, social, legal, and financial risks is found to be positively associated with online addiction (Sugaya et al.,
2019). Cognitive factors associated with IGD include maladaptive beliefs, such as the excessive value attributed to gaming, biases in gaming, over-reliance on gaming to meeting their emotional needs and the need for gaming to attain social acceptance (King & Delfabbro, 2016). Lastly, family factors have been found to play a significant role in the etiology and development of IGD
(Bonnaire & Phan, 2017). In families with better cohesion, structure, functional connectivity and less conflict, the risk of problematic gaming is significantly decreased (Bonnaire & Phan, 2017).
Recently, the connection between attachment security and IGD has been increasingly explored, with several studies suggesting a positive correlation between insecure attachment and IGD.
However, further work is needed in order to improve our understanding of the behavioral pathways through which attachment contributes to the etiology and development of ATTACHMENT SECURITY AND INTERNET GAMING DISORDER 11 psychopathology. The purpose of this study is to further explore the link between attachment security and IGD.
Attachment Theory
Attachment and early life unpredictability play a significant role in the development of children and can provide much insight into the social development of individuals well into adulthood (Szepsenwol, Simpson, Griskevicius & Raby, 2015). Researchers recently have begun to understand the implications of attachment and attachment security in the development of psychopathologies such as gaming disorder (Throuvala et al., 2019; Eichenberg, Schott, Decker
& Sindelar, 2017; Danet & Mijkovitch, 2017). Research in attachment has provided much insight into individuals’ vulnerability to stress-factors, decreased adaptability, decreased social self- efficacy and impaired decision making associated with insecure and anxious patterns of attachment (Christian, Sellbom. & Wilkinson, 2019).
According to Bowlby (1988) and Ainsworth’s Theory of Attachment (1969), the initial relationship between a child and their parent determines the quality and type of attachment bond that a given child forms, which in turn influences the child’s cognitive, social and psychological development (Bretherton, 2015). One’s early childhood relationships with caregivers shape the child’s schemas of relations that translate into their expectations in adolescence as well as into adulthood. It informs them on how to interact socially with their peers, friends and partners
(Bretherton, 2015; Smuts, 1985). These schemas are referred to by Bowlby (1988) as working models that one develops for oneself and others. They develop from a series of interpersonal interaction patterns. If a child does not receive comfort, protection or respect for their needs, then they develop unworthy and incompetent self-models (Bretherton, 2015). If the child does receive ATTACHMENT SECURITY AND INTERNET GAMING DISORDER 12 comfort and security, then they form a valued and self-reliant working model (Bretherton, 2015;
Smuts, 1985).
Children then use these working models to predict their attachment figures’ behavior and use them in developing their own response patterns (Bretherton, 2015). According to Bowlby
(1988), stable and self-reliant adults usually have a balanced relationship with their parents that offers them both support and autonomy when needed. This translates to their own personal behaviors in interpersonal relationships, including that with their own children, thus creating an intergenerational transmission pattern of attachment (Bowlby, 1988; Snyder, 2011). In a meta- analysis exploring the extent to which early attachment patterns affect adult attachment behaviors, Fraley (2002) found that representations of early attachment experiences maintain a continued influence on attachment behaviors well into adulthood. It is also argued that these working models of relationships are subject to change throughout life due to new experiences
(Fraley, 2002). However, the carryover effects of early attachment remain strong throughout adulthood (Smuts, 1985). This provides support for the assessment of attachment in adults rather than assessing parental relationships.
Attachment Categories
The initial relationship between a child and their caregiver results in the formation of one of four types of attachment: secure, anxious-preoccupied, fearful-avoidant and dismissive- avoidant (Bowlby 1988; Snyder, 2011). Similarly, in adults, attachment orientation can be categorized as secure, insecure-anxious or insecure-avoidant (Fraley, Waller & Brennan, 2002).
In secure attachment, the initial relationship between a child and caregiver is characterized by emotions of safety, empathy, mutually shared pleasure and positive affect (Bowbly,1988; Siegel
& Hartzell, 2003). This translates to stable and secure relationships in adulthood as well. ATTACHMENT SECURITY AND INTERNET GAMING DISORDER 13
Insecure attachment in childhood can translate into two types of attachment styles: anxious and avoidant attachment (Bowlby, 1988; Fraley, 2002). In anxious attachment, individuals and children feel a level of anxiety or fear of rejection in interpersonal relationships though they may crave connection; this anxiety limits them in interpersonal relationships. In avoidant attachment, children and adults experience discomfort when making emotional connections and attempt to create distance between themselves and others (Snyder, 2011; Bowlby, 1988; Fraley, 2002). This could either be due to fear (fearful-avoidant) or an inability to emotionally connect (dismissive- avoidant). Avoidant individuals tend to be less optimistic, less cheerful and have greater negative affect. Often they experience difficulty in forming bonds and romantic relationships, and this insecurity in attachment is also translated into their own attachment bond with their children
(Bowlby, 1988).
Attachment and IGD
A significant motivator in the use of the internet is the pursuit of social connection and fulfillment of emotional needs (Danet & Miljkovitch, 2017). Recently, researchers have sought to explore the possibility that attachment style may lead to an overreliance on the internet as a means of finding connection and in-turn lead to problematic internet use (Danet & Miljkovitch,
2017). Danet & Miljkovitch (2017) posited that the link between attachment and gaming addiction is due to the safe space for self-expression offered by the internet for insecurely attached individuals. It allows them to connect with others while maintaining a distance from others physically, mentally and emotionally (Danet & Miljkovitch, 2017). Study results indicated that insecurely attached individuals are more likely to have issues with excessive internet use as compared to their securely attached peers (Danet & Miljkovitch, 2017). ATTACHMENT SECURITY AND INTERNET GAMING DISORDER 14
Furthermore, studies have indicated a significant positive correlation between problematic gaming, internet addiction and insecure attachment (Eichenberg et al., 2017;
Tavakoli, Jomehri & Farrokhi, 2014). For example, it has been demonstrated by Eichenberg et al. (2017) that self-reported measures of internet addiction, using the scale for online addiction for adults, positively correlated with self-reported measures of insecure attachment using a questionnaire measuring partnership expectations. This suggests that there may be a potential relationship between insecure attachment and problematic internet use.
Furthermore, in a study assessing patients in an internet addiction outpatient clinic,
Şenormancı, Güçlü & Konkan, R. (2014) demonstrated that there is a significant positive relationship between internet addiction tendencies and self-reported anxious attachment scores.
This suggests that perhaps anxious attachment should be assessed more closely when determining the relationship between attachment and IGD. The study had a very small participant pool of 30, leaving room for further studies to be conducted with larger sample sizes.
For example, Shin, Kim & Jang (2011) used a larger sample size when assessing anxious and avoidant attachment as potential predictors of problematic internet-use (in alcohol addiction patients) and found a significant positive relationship between internet addiction and anxious attachment, as well as avoidant attachment.
Though several studies have assessed the relationship between attachment and internet addiction, few have assessed insecure attachment in IGD. However, of those that have, results have indicated that attachment may play a key role in the etiology of IGD. For example,
Throuvala et al. (2019) assessed self-reported parental rejection and IGD scores and found a significant positive relationship between IGD and parental rejection. This suggests that the parental relationship may have a role in IGD. Furthermore, in a case study on two teenage boys ATTACHMENT SECURITY AND INTERNET GAMING DISORDER 15 with severe IGD, Benarous et al. (2019) found that both boys grew up in adverse environments, which resulted in insecure attachment styles and a high degree of difficulty forming relationships. Following an adverse life-event, both boys began gaming excessively and demonstrating compulsive internet abuse. These findings suggest that early attachment may play a major role in increasing vulnerability to development of psychopathology. Further research is needed, however, in order to understand the pathways through which insecure attachment leads to IGD. One way through which we can further our understanding of the relationship between attachment and IGD is through the study of stress and stress susceptibility in insecure attachment and IGD.
Attachment and Susceptibility to Stress
Stress-vulnerability is a potential consequence of early experiences of trauma which are significantly correlated to insecure attachment. Before delving into the discussion of the relationship between stress-vulnerability and insecure attachment, it is important to note that evidence supporting this relationship may be limited in that they are indicative of correlation and not causation. The existing literature has demonstrated potential irregularities in cortisol reactivity and coping ability in both insecurely attached children and adults (Sroufe & Waters,
1977; Hill-Soderlund et al., 2008; Gander & Buchheim, 2015; Gunnar, Brodersen, Nachmias,
Buss & Rigatuso, 1996). For instance, Gunnar et al. (1996) demonstrated that insecurely attached infants showed significantly higher cortisol reactivity in response to Ainsworth’s strange situation test in comparison to their securely attached counterparts. This suggests that insecure attachment may result in a greater susceptibility to stress. Furthermore, Nachmias, Debuse &
Powers (2013) found that attachment shapes both an individual’s perceptions of stress as well as ATTACHMENT SECURITY AND INTERNET GAMING DISORDER 16 their cortisol response to stress. An association between cortisol response and attachment style was also noted in Pietromonaco et al.’s (2013) experiment. These findings suggested that HPA activation may be an integral part of the biological pathways that relate attachment and health outcomes.
Communication between mother and infant is an essential component in the development of secure attachment (Schore, 2003). When the affective behavior of a mother and child are synchronized, there is a regulation of positive arousal and a healthy pattern of engagement and disengagement between the child and the mother. This is a vital component of the healthy affective development of children and results in development of healthy coping strategies, in turn leading to secure attachment (Schore, 2003). When the interactions and affective behaviors between mother and child are asynchronized, such as in cases where the parent is abusive or detached, the child becomes insecurely attached and develops negative coping strategies as well as a negative sense of self (Schore, 2003). For example, in insecurely attached children, when subjected to stress, behaviors such as screaming, crying and vomiting are observed (Schore,
2003). When the mother or parent is non-reactive or does not provide the child with comfort, the child begins to disengage from the external world exhibiting dissociative coping strategies
(Kaufman & Rosenblum, 1967), These traumatic experiences in childhood have profound effects on the early behavioral and psychological development of children (Schore, 2003).
Stress Susceptibility in Insecure Adults
The effects of early trauma result in deficits in adaptive functioning and processing of emotional stimuli (Schore, 2003). Studies have demonstrated that these effects can persist well into adulthood (Raine et al., 2001; Schore, 2003). For instance, Raine et al. (2001) demonstrated ATTACHMENT SECURITY AND INTERNET GAMING DISORDER 17 that adults who were severely abused during their childhood, and developed PTSD later on, have reduced right-hemisphere activation, specifically, in the right temporal cortex during a visual and verbal working memory task. This lends to potential evidence suggesting that the consequences of early life unpredictability and insecure attachment may be far-reaching well into adulthood.
Furthermore, similar findings to Gunnar et al. (1996) in children have also been observed in adults. Pietromanaco et al. (2013) found that insecurely attached women show greater cortisol reactivity in comparison to their securely attached counterparts. This further supports the notion that the neurophysiological effects of attachment are significant even in adulthood. Furthermore,
Muller, Sicoli & Lemieux (2000) found that of the individuals assessed (victims of childhood abuse), who also reported posttraumatic stress symptoms, 76% were found to have an insecure attachment style. Additionally, Bessar, Neria & Haynes (2009) found that in Israeli citizens (of whom were exposed to terrorist attacks), those with attachment anxiety who perceived a greater level of stress showed more symptoms of PTSD as compared to those with no attachment anxiety. Attachment anxiety was also found to be related to greater vulnerability to development of PTSD (Bessar et al., 2009).
The literature cited above is indicative of potential support for the relationship between attachment and susceptibility to stress. However, since much of the research indicated above is correlational in nature, further work is needed in order to establish the relationship between attachment and stress susceptibility. It can be argued that observed changes in stress response such as increases in cortisol may be directly related to the experience of trauma itself. That experience of trauma and irregulating of the stress system may then, in turn, lead to a development of insecure attachment. Additionally, it should also be noted that in adults, studies have found blunted cortisol response in response to trauma such as PTSD. For example, Yehuda ATTACHMENT SECURITY AND INTERNET GAMING DISORDER 18
(1998) posited that in studies assessing adults with PTSD, there is an observed blunting of the adrenocorticotropic hormone to corticotrophin-releasing factor. Corticotrophin-releasing factor is released as a result of stress exposure. Adrenocorticotropic hormone, in turn, is responsible for cortisol release.
IGD and Susceptibility to Stress
While research on IGD and stress susceptibility is limited, vulnerability to stress and difficulty dealing with stressful situations have also been observed in individuals that showcase
IGD. For example, youth with IGD have been found to showcase a heightened vulnerability to stress in comparison to healthy controls (Kaess et al., 2017). Youth with IGD reported greater levels of perceived stress on a daily basis, and showed attenuated cortisol response in Kaess et al.’s study assessing stress vulnerability in youth with IGD. This suggests that that the stress response system may be altered in those with IGD. One way through which one can understand the relationship between IGD, and stress vulnerability is through the Interaction of Person-
Affect-Cognition-Execution (I-PACE) model of specific internet-use disorders (Brand et al.,
2016). The I-PACE model suggests that stress vulnerability is a major pre-disposing factor in
IGD. Stress-reduction is a major motivator amongst problematic gamers who use gaming as a coping mechanism (Kaess et al., 2017).
In a sense, it is a form of a dissociative coping mechanism seen in adolescents and adults.
Though few studies have been done on IGD and stress response, there is also evidence to suggest that deficits in stress response are common in behavioral addiction. For example, in an assessment of patients with gambling disorder Wohl et al. (2008) found elevated early morning cortisol levels in pathological gamblers as compared to recreational gamblers. This suggests a ATTACHMENT SECURITY AND INTERNET GAMING DISORDER 19 potential relationship between stress susceptibility and behavioral addiction. Further research is needed to improve our understanding of the relationship between stress susceptibility and IGD.
The Current Study
On the basis of the literature reviewed, the aim of this study was to investigate the relationship between Attachment and Internet Gaming Disorder. Although past studies have shown significant positive correlations between insecure attachment and Internet Gaming
Disorder, no studies to date had assessed the extent to which vulnerability to stress impact this relationship (Eichenberg et al., 2017; Tavakoli, Jomehri & Farrokhi, 2014). In addition, past research has indicated that there is a significant relationship between both attachment and stress susceptibility and IGD and stress susceptibility. Therefore, this study sought to further investigate the significance of this relationship by assessing the potential mediating effect of stress susceptibility on the relationship between attachment and IGD.
Hypotheses and Predictions
In concordance with the previous literature, the current study hypothesized that in comparison with securely attached participants, insecurely attached individuals would show greater levels of excessive gaming as measured by the Internet Gaming Disorder Scale (IGDS;
Lemmens, Valkenburg & Gentile, 2015; Danet & Miljkovitch, 2017; Şenormancı et al., 2014;
Eichenberg et al., 2017; Tavakoli, Jomehri & Farrokhi, 2014). It was also hypothesized that
Insecurely attached individuals would exhibit higher perceived stress scores as measured by the
Stress Overload Scale (SOS; Amirkhan, Urizar & Clark, 2015; Sroufe & Waters, 1977; Hill-
Soderlund et al., 2008; Gander & Buchheim, 2015; Gunnar et al., 1996; Schore, 2003). There ATTACHMENT SECURITY AND INTERNET GAMING DISORDER 20 were no hypotheses pertaining to the differences between the different categories of attachment, except for avoidant and anxious attachment. A significant positive relationship is predicted between avoidant and anxious subscale scores on the ECR scale and IGD, insecure attachment and stress susceptibility, and IGD and stress susceptibility (Kaess et al., 2017). Lastly, it was predicted that stress susceptibility, will mediate the relationship between anxious & avoidant attachment and IGD.
Methods
Study
In this study, adults 18 and older were assessed for internet addiction tendencies as well as IGD. Data was collected through both an online research platform (MTURK), as well as
SONA at the University of Minnesota, Duluth. Participants were assessed for internet gaming addiction tendencies as well as their attachment orientations using the Internet Gaming Disorder
Scale (IGDS; Lemmens, Valkenburg & Gentile, 2015) and the Experiences in Close
Relationships Scale-Revised (RSQ-R: Brennen, Clark & Shaver, 1998) respectively. Participants were also asked to complete a questionnaire in which they reported on susceptibility to stress, and demographics measures such as the Adverse Childhood Experiences Questionnaire (ACE) trauma self-report questionnaire.
Population
In this study, the participant pool consisted of individuals with a minimum age of 18. The sample consisted of both college students and/ or online MTURK participants (N = 482). A ATTACHMENT SECURITY AND INTERNET GAMING DISORDER 21 power analysis was performed in order to determine the necessary sample size prior to completion of the study. This determined that for the purpose of this study, the ANOVA analysis requires at least 176 participants, and the multiple regression analyses require an additional 189 participants. This was calculated using the Heinrich sample size calculator which indicated that with the given confidence interval (95%) and accepted effect size (.05), a sample size of at least
365 is required for the study (Heinrich, 2019).
To control for the external influences of education, participants were required to have at least 12 years of schooling. All participants were also required to have used the internet and played an internet game at least once in the past year, and were also required to be proficient in
English. The study was advertised as an attachment and gaming study. Prior to the study, all materials and procedures were approved by the Institutional Review Board at the University of
Minnesota. Informed consent was obtained from all participants prior to completion of assessments.
In the initial data collection process, data were collected from 482 participants in total.
Two attention checks were included in the surveys and used to exclude participants inattentive response patterns (n = 43). Participants were also excluded for failing to report when they had last played a video game (n =15). Individuals were excluded for this largely because this was essential to our research question. Lastly, participants were also excluded for being under the age of 18 (n =1) . The final population was comprised of n = 423 participants, with 13.5% of the data excluded (n = 59).
Participants were all screened for inclusion and exclusion criteria before and during the assessment. Exclusion criteria included: All individuals who did not use the internet or have not played an internet game in the last year (or failed to respond to this demographic question), any ATTACHMENT SECURITY AND INTERNET GAMING DISORDER 22 individuals who were not fluent in English (so that discrepancy in scores due to a language barrier were avoided), individuals under 18, and individuals with less than 12 years of schooling.
Measures
The Internet Gaming Disorder Scale. Developed by Lemmens, Valkenburg & Gentile
(2015), the IGD Scale is a measure of DSM-5 criteria for Internet Gaming Disorder (IGD).
Criteria for IGD include preoccupation with gaming, increased tolerance to time spent gaming, perpetual desire to game, the use of gaming as a coping mechanism, experiencing negative consequences as a result of gaming behavior, conflict related to gaming, deception of others in order to hide gaming behaviors, and the domination of gaming behaviors overall social and recreational activities (American Psychiatric Association, 2013). In this study, the IGDS was used as a measure of gaming addiction. Its criteria are in line with the criteria suggested by the
APA in the Diagnostic and Statistical Manual of Mental Disorders (DSM-5; APA, 2013). The
IGDS is a dichotomous scale consisting of 27 items for which individuals answer ‘yes’ or ‘no’
(yes = ‘this applies to me’, no = ‘this does not apply to me’). Items include questions pertaining to the frequency of gaming, consequences experienced gaming, difficulty experienced in the cessation of gaming and more. Scores on the IGDS can range from 0 to 27. Higher scores indicate greater levels of disorder. At least 5 of the criteria must be met in order to be classified as a disordered gamer. The IGDS has good internal consistency and reliability (α = 0.93).
However, an error was made in entering the scale into Qualtrics where one item was left out. As a result, a new Cronbach’s alpha of α = .93 was calculated for this scale, indicating that the scale still had good internal consistency and reliability. ATTACHMENT SECURITY AND INTERNET GAMING DISORDER 23
Experiences in Close Relationships Scale – Revised. The Experiences in Close
Relationships Scale -Revised (ECR- R; Fraley, Waller & Brennan, 2002) is a revised version of the original experiences in close relationships scale by Brennen, Clark & Shaver (1998). This scale is a self-report questionnaire that detects individual differences in people and how they approach relationships. It evaluates attachment-related anxiety and attachment-related avoidance.
The purpose of this questionnaire is to help determine whether those with more insecure and avoidant attachment styles showcase a greater extent of internet addiction and gaming disorder
(ECR -R; Fraley, Waller & Brennan, 2002). This questionnaire was used to assess whether individuals had a secure attachment, dismissive-avoidant, anxious-avoidant or preoccupied styles of attachment. The ECR is considered a valid measure of attachment and is high in validity and reliability (Kurdek, 2002). The revised version of the ECR by Fraley, Waller & Brennan (2002) was used in this study. This revised measure assesses attachment on two subscales: Anxiety and
Avoidance. Avoidant individuals are defined as those who seek independence and avoid intimacy, and anxious individuals are defined as those who avoid intimacy due to a fear of rejection (Fraley et al., 2002). The ECR-R is a 36-item measure of adult attachment. It is a measure of how individuals generally experience relationships. It is different from other forms of attachment scales in that it does not just focus on a specific partners or individuals such as a parent. This scale is scored on a Likert scale of 1 to 7 (1 = ‘Strongly Disagree’, 7 = ‘Strongly
Agree’). The first half of the scale consists of items concerning attachment related anxiety (Items
1 – 18). The second half consists of items concerning attachment related avoidance (Items 19 –
36). The average score of items 1-18 is a measure of attachment anxiety. The average score of items 19-36 is a measure of attachment avoidance. In order to obtain a score however, items 9 and 11 were first be reverse scored. Following scoring, participants were assigned to four ATTACHMENT SECURITY AND INTERNET GAMING DISORDER 24 quadrants depending on their scores on the avoidant and anxiety subscale scores on the ECR scale. Individuals with low anxiety and high avoidance were categorized as dismissive-avoidant.
Those with low anxiety and low avoidance were categorized as secure, those with high anxiety and low avoidance were categorized as preoccupied, and those with high anxiety and high avoidance were categorized as fearful-avoidant. For regression analyses, all groups other than secure group (n = 109) were grouped under the insecure category (Preoccupied, Dismissive-
Avoidant & Fearful Avoidant; n = 314). A Cronbach’s alpha was calculated for this scale indicating high internal consistency, convergent validity and reliability (α = .938).
The Stress Overload Scale. The Stress Overload Scale (SOS) is a measure of stress susceptibility in adults which assesses individual’s vulnerability to overwhelming demands
(Amirkhan, Urizar & Clark, 2015). The SOS contains two subscales measuring personal vulnerability (PV) to stress and event load (EL). Even numbered items pertain to event load, while odd numbered items pertain to personal vulnerability when reverse scored. In addition, the scale includes 6 items used to discourage a pattern of negative responding. Items on the SOS are scored on a 5-point Likert scale (0 = ‘Not at all’, 5 = ‘A lot’). All items on the scale are summed to obtain a continuous total score. Higher scores are indicative of greater levels of stress overload. Subscales can be split into our different categories including high EL- high PV, low
EL-low PV, high EL-low PV and low EL-high PV. Those categorized as both high EL and high
PV are at the greatest risk for pathology. For the purpose of this study, only the continuous total score was used. The SOS has high internal consistency (α > .94) as well as good test-retest reliability (r = .5). Additionally, studies have also demonstrated the SOS’s ability to predict abnormal cortisol reactivity to stressful events (Amirkhan et al., 2015). A new Cronbach’s alpha was calculated for this scale indicating high internal consistency and reliability (α = .96). ATTACHMENT SECURITY AND INTERNET GAMING DISORDER 25
Adverse Childhood Experiences Questionnaire. The Adverse Childhood Experiences
Questionnaire (ACE; Felitti et al., 1998) is an adult measure for the experiences of physical, psychological and sexual abuse in childhood. Childhood is specifically defined as the first 18 years of life. It also includes four categories measuring household dysfunction such as substance abuse, parental abuse, mental illness and criminal behavior of household members. In total, the
ACE includes 10 questions for which participants respond ‘yes’= (I have experienced this) or
‘no’= (I have not experienced this). Total scores were determined by the sum of the individual categories. Scores on the ACE range from 0 to 7 (0 = ‘unexposed’, 7 = ‘exposed to all categories’). This scale was used to determine whether experiences of early life trauma covariate with attachment orientations as well as other measures included in the study. Measuring trauma is important because it can be argued that the effects and changes in cortisol seen in insecure attachment, are a result of trauma, not necessarily insecure attachment itself. Studies assessing the validity of the ACE questionnaire have found the ACE to be reliable and appropriate for research purposes (Bethell, Carle, Hudziak, Gombojav, Powers, Wade & Braveman, 2017). A
Cronbach’s alpha was calculated for this scale which indicated high internal consistency and reliability (α = .88).
Procedure
The current study was completed using two data collection platforms. The first was
SONA, a student research experience program run by the University of Minnesota- Duluth that is used primarily for studies within the psychology department. Students who participated through
SONA voluntarily signed up for the study on the platform and completed the surveys after electronically signing informed consent forms. They then completed the demographic measures ATTACHMENT SECURITY AND INTERNET GAMING DISORDER 26 followed by the survey’s listed and described above. Following completion of the study, students were rewarded 1 research experience point. The second platform utilized was MTURK or
Amazons Mechanical Turk. The major of data collection was done on this platform. This platform is a public platform used by researchers and companies in order to collect data from the general public, or MTURK users, from around the world. Participants on MTURK were screened based off of the study requirements. If they met the criteria for the study, they then completed the informed consent and completed the study. Participants on MTURK were compensated for their participation with a payment of $2.00. All participants accessed the survey through the Qualtrics platform and were provided with debriefing forms at the end of the survey. Procedures were approved by the relevant institutional research ethics board.
Statistical Analysis
The hypothesis that gaming addiction symptoms would differ by attachment group was analyzed with a one-way ANOVA with attachment as the grouping factor. The dependent variable was analyzed using the IGDS scale. Following this, another one-way ANOVA was used to assess if significant differences existed between attachment groups (Secure, dismissive- avoidant, Fearful Avoidant and preoccupied) in perceived stressSignificant ANOVA results were followed with Tukey’s HSD post hoc analyses to assess differences between attachment groups on IGD symptoms, and perceived stress respectively. Additionally, since errors were made in the
IGDS measure in which one question was not presented to the participants, Cronbach’s alphas were calculated for all measures including the IGDS to assess for convergent validity and reliability of measures. Means and standard deviations were calculated for all variables concerning the study. ATTACHMENT SECURITY AND INTERNET GAMING DISORDER 27
Following the ANOVA analyses, hierarchical regression analyses were completed to assess the relationships between attachment and Internet gaming addiction symptoms. The secure
(n = 109) and insecure attachment (n = 314) groups were compared in this analysis (Insecure = 0,
Secure = 1). Additionally, zero-order correlations were assessed with a bi-variate correlation matrix to ensure that no multicollinearity existed between the independent variables. Lastly, standardized z-scores were assessed to ensure that no multivariate outliers existed in the data. A hierarchical regression analysis with 4 steps was completed to assess the hypothesis predicting a significant positive relationship between insecure attachment styles and IGD symptoms while controlling for demographic variables including sex and age. Additionally, a hierarchical regression analysis with 4 steps was also completed to assess the hypothesis predicting a significant positive relationship between insecure attachment styles and perceived stress while controlling for demographics. Lastly, A hierarchical regression analysis with 4 steps was also completed to assess the hypothesis predicting a significant positive relationship between perceived stress and gaming addiction scores. Hierarchical regression analyses controlled for both age and sex (female = 0, male = 1). Prior to this, histograms for each measure were assessed for normality and skew.
Following statistically significant results in the hierarchical regressions, mediation analyses were completed to assess the hypotheses predicting that perceived stress or stress susceptibility will mediate the relationship between attachment and IGD symptoms.
Results
Demographic characteristics were assessed for participants in this study. The final sample of participants was primarily male (69.9%), Caucasian (54.1%), with a mean age of 30.48 (SD = ATTACHMENT SECURITY AND INTERNET GAMING DISORDER 28
8.57, range = 18-64). Males (M = 13.69, SD = 6.97) showed higher scores than females (M =
8.72, SD = 7.31) on gaming addiction tendencies, t(420) = -6.63, p <.001. Descriptive statistics for gaming disorder scores, perceived stress, trauma and demographics are included in Table 1, calculated separately by attachment style. Participants were categorized into 4 attachment groups based on attachment related avoidance and attachment related anxiety subscale scores on the
ECR-R assessment. The majority of participants were categorized as having preoccupied styles of attachment (n = 144; 34.04%). 109 participants were categorized as secure (25.77%), 41 were categorized as Dismissive-Avoidant (9.69%) and 129 were categorized as Fearful-Avoidant
(30.50%). Thus 109 were categorized as secure (25.77%), and 314 were categorized as insecure
(74.23%). This is atypical, as in most studies, typically 60% of individuals are categorized as having secure attachment (Rees, 2007). However, more recent research has indicated that the proportion of insecure attachment styles has increased to 60% over the years in student populations (Konrath et al., 2014).
Additionally, there were significant differences between attachment styles in trauma scores, F(3,419) = 68.96, p <.001. A Tukey’s post-hoc analysis showed that while the secure group did not differ significantly from the Dismissive group (p = ns.), it did differ significantly from the preoccupied group (p <.001), and the Anxious-Avoidant group (p <.001). Individuals with preoccupied styles of attachment (M = 6.76, SD = 2.94) reported higher trauma scores than individuals with secure styles of attachment (M = 1.76, SD = 2.11), individuals with Anxious–
Avoidant styles of attachment (M = 4.50, SD = 3.32), and individuals with Dismissive-Avoidant styles of attachment (M= 2.51, SD = 2.83). ATTACHMENT SECURITY AND INTERNET GAMING DISORDER 29
Attachment styles and internet gaming addiction
Differences between attachment groups were assessed using a one-way ANOVA with attachment styles (Secure, Dismissive-Avoidant, Fearful-Avoidant & Preoccupied) as the independent variable and gaming addiction scores as the dependent variable. Significant effects of attachment type were found on Internet gaming disorder scores (F (3,419) = 70.80, p < .001, w = .33). A post-hoc analysis was completed with a Tukey’s post-hoc analysis showied that while the secure group did not differ significantly from the dismissive-avoidant group (p = ns), it did differ significantly from the Anxious-Avoidant group (p < .001), and Preoccupied group (p <
.001). Additionally, between-group differences were found between the Dismissive-Avoidant group, Anxious-Avoidant group, and Preoccupied group at p < .001. In accordance with the hypothesis that insecurely attached participants will show greater levels of excessive gaming than securely attached individuals, secure individuals (M=6.13, SD = 4.75) showcased much lower gaming addiction scores than Anxious-Avoidant individuals (M = 13.23, SD = 6.33), with
Preoccupied individuals (M= 16.94, SD= 7.45) exhibiting to highest levels of gaming addiction.
Additionally, Dismissive individuals (M= 8.76, SD= 1.06) exhibited much lower gaming addiction scores than Anxious-Avoidant individuals and Preoccupied individuals.
Attachment Styles and Perceived stress
Differences between attachment groups were also assessed using a one-way ANOVA with attachment styles (Secure, Dismissive-Avoidant, Fearful-Avoidant & Preoccupied) as the independent variable and perceived stress as the dependent variable. Significant effects of attachment type were found on SOS scores (F(3,419) = 126.56, p < .001, w = .46). A Tukey’s post-hoc analysis showed that the secure group differed significantly from the preoccupied group
(p <.001), and the Anxious-Avoidant group (p <.001), but not the Dismissive group (p = ns.). In ATTACHMENT SECURITY AND INTERNET GAMING DISORDER 30 accordance with the hypothesis that insecurely attached participants will show greater levels of perceived stress, secure individuals (M=51.21, SD = 23.90) showcased much lower levels of perceived stress than Anxious-Avoidant individuals (M = 86.19, SD = 24.07), with Preoccupied
Individuals (M= 110.01, SD= 26.62) exhibiting to highest levels of perceived stress. A post-hoc analysis was completed with Tukey’s test showing that while the secure group did not differ significantly from the dismissive-avoidant group (p = ns), it did differ significantly from the
Anxious-Avoidant group (p < .001), and Preoccupied group (p < .001). Additionally, between- group differences were found between the Dismissive-Avoidant group, Anxious-Avoidant group, and Preoccupied group at p < .001, with Dismissive individuals (M= 8.76, SD= 1.06) exhibiting lower levels of perceived stress than Anxious-Avoidant individuals and Preoccupied individuals.
Zero-Order Correlations
Bivariate correlations were calculated between IGD scores, SOS scores, and the anxious and avoidant subscales scores of the ECR (see Table 2). Although data was collected using the
Internet Addiction Test as well as the Compulsive Internet Use Scale, only data pertaining to the
Internet Gaming Disorder Scale (IGDS) as the literature indicates that research is moving towards the IGDS as the preferred method for assessment of gaming addiction. Additionally, data was collected using both the Perceived Stress Test (PSS) and SOS perceived stress scales.
For the purpose of this analysis, only data for the SOS was included as it is well supported by the literature, and according to reliability analyses, has greater convergent validity (α = .96). As expected, IGDS scores were significantly associated with Attachment Anxiety (r = .63, p < .01),
Attachment Avoidance (r = .26, p < .01), and ACE trauma scores (r = .69, p <.01). Gaming addiction scores were also found to be significantly related to the last time gamed (r = .11, p <
.05). ATTACHMENT SECURITY AND INTERNET GAMING DISORDER 31
Hierarchical Regression Analyses
Refer to Figure 1 for the mediation model tested in these analyses. In order to assess hypotheses pertaining to the relationship between attachment and gaming addiction scores, and the potential mediation through perceived stress, a hierarchical regression was conducted in order to test the pathways in the mediation model following the Barron & Kenny (1986) method for testing mediation through regression (see Tables 3-6). The first hierarchical regression analysis assessed whether internet gaming addiction scores could be predicted from attachment orientation (secure vs. insecure) (see Table 3). In the first step, demographic variables were entered to control for gender and age. The hierarchical multiple regression revealed that at step one, Sex and Gender contributed significantly to the regression model, (F (2,419) = 26.51, p <
.001) and accounted for 10.8% of the variation in IGDS Scores. Men had significantly higher gaming addiction tendencies than women (β = 5.28 p < .001). Additionally, as age increased, so did internet gaming addiction scores (β = .12, p = .004). In the second step, introducing the attachment variables explained an additional 29.8% of variation in IGDS Scores and this change in R² was significant, F (3,418) = 60.53, p < .001. As predicted, attachment category (β = -7.50, p < .001) was significantly associated with gaming addiction scores above and beyond the effects of the demographic variables.
Table 4 provides results for the second hierarchical regression completed to determine if perceived stress could be predicted from attachment orientation (secure vs. insecure). In the first step, demographic variables were entered to control for gender and age. The hierarchical multiple regression revealed that in step 1, Sex and Gender contributed significantly to the regression model,(F (2,419) = 8.34, p < .001) and accounted for 3.4 % of the variation in SOS
Scores. Men reported significantly higher levels of stress than women (β = 14.69, p < .001). ATTACHMENT SECURITY AND INTERNET GAMING DISORDER 32
Next, the Attachment variable was introduced into the model in step 2, explaining an additional
30% of variation in SOS Scores and this change in R² was significant, F (2,418) = 61.61, p <
.001. As predicted, attachment category (β = -41.08, p < .001) was significantly associated with perceived stress scores above and beyond the effects of the demographic variables. When accounting for attachment, men still reported significantly higher levels of stress than women (β
= 8.08, p = .01).
Next, Table 5 depicts results for the third hierarchical regression analysis completed to determine if perceived stress is predictive of gaming addiction scores. In the first step, demographic variables were entered to control for gender and age. The hierarchical multiple regression revealed that at stage one, Sex and Gender contributed significantly to the regression model, (F(2,419) = 26.51, p < .001) and accounted for 10.8% of the variation in IGDS Scores.
Men reported significantly higher levels of stress than women (β = 5.28, p < .001). Additionally, as age increased, so did levels of perceived stress (β = .12, p= .01). In the next step, introducing the SOS Stress variable explained an additional 43.9% of variation in IGDS Scores and this change in R² was significant, F(3,418) = 110.65, p < .001. As predicted, perceived stress (β =
.13, p < .001) was significantly associated with gaming addiction scores above and beyond the effects of the demographic variables. Men continued to report significantly higher levels of stress than women (β = 3.41, p < .001).
Lastly, the overall model was tested using a hierarchical regression analysis to determine if attachment and perceived stress together significantly predicted gaming addiction (see Table
6). In the first step, demographic variables were entered to control for gender and age. The hierarchical multiple regression revealed that at stage one, Sex and Gender contributed significantly to the regression model, F (2,419) = 26.51, p< .001) and accounted for 10.8% of the ATTACHMENT SECURITY AND INTERNET GAMING DISORDER 33 variation in IGDS Scores. Next, Introducing the SOS Stress variable explained an additional
43.9% of variation in IGDS Scores and this change in R² was significant, F (3,418) = 110.65, p <
.001.
Lastly, introducing the attachment variable explained an additional 46.2% of variation in
IGDS Scores and this change in R² was significant, F (4,417) = 91.40, p < .001. As predicted, perceived stress (β = .11, p < .001) and attachment (β = -3.17, p <.001) were significantly associated with gaming addiction scores above and beyond the effects of the demographic variables. When controlling for attachment, perceived stress remained significant indicating that some form of mediation is supported. Additionally, both attachment and perceived stress significantly predicted gaming addiction scores indicating partial mediation. These results were followed up with a PROCESS mediation analysis to confirm that the mediation model presented in this paper is truly significant and that findings are not a result of family-wise errors.
Mediation Analysis
Hierarchical regression analyses were completed to assess the hypothesis that perceived stress mediates the effect of attachment security on gaming addiction scores. Results supported the hypothesis that attachment is a significant predictor of perceived stress (β = -41.08, SE =
3.23, p < .001), and that perceived stress in turn predicted gaming addiction scores above and beyond the covariates, (β = .11, SE = .01, p < .001). The regression analysis provided support for the mediational hypothesis. While controlling for the mediator, attachment type was still a significant predictor of gaming addiction scores, b = -3.17, SE = .72, consistent with partial mediation. About 46.7% of the variance in the outcome variable, gaming addiction, was accounted for by the antecedent variables while covariates (sex and age) were held constant (F
(3,417) = 91.40, p < .001). Utilizing bootstrap estimates of 5000 samples, the indirect effect was ATTACHMENT SECURITY AND INTERNET GAMING DISORDER 34 assessed using the PROCESS macro Version 3.4 (Hayes, 2017). The indirect effect of attachment on gaming addiction was significant above and beyond the covariates, β = -.58, 95%
CI [-0.720, -0.458]. Individuals with secure attachment scored -.58 points lower than insecure individuals on gaming addiction as mediated by perceived stress.
Discussion
Gaming addiction, a relatively new topic in research, has attracted much interest from the clinical research community in recent years. Internet gaming was recently added as a tentative disorder in the DSM-5 and classified by the World Health Organization as an official disorder in the ICD-11 (Kaess et al., 2017; WHO, 2018). The negative consequences associated with gaming addiction run almost parallel to those seen with substance abuse. Symptoms include excessive time investment, experience of anxiety, depression, denial and more (APA, 2013;
Griffiths, 2005; Monacis et al., 2017). While there is much research on the consequences of excessive gaming, there is still much left to discover about the etiology of IGD. The analysis of literature revealed potential relationships between attachment security and gaming addiction, with many studies suggesting positive correlations between insecure attachment and gaming addiction as well as stress susceptibility and gaming addiction (Eichenberg et al., 2017; Danet &
Miljkovitch, 2017; Şenormancı et al., 2014; Gander & Buchheim, 2015).
This study attempts to expand on current literature regarding the pathways through which attachment may influence the onset and progression of gaming addiction. This was done through the assessment of the relationship between both attachment and stress susceptibility and IGD and stress susceptibility. Furthermore, this study sought to expand on recent findings through the assessment of the potential mediating effects of stress susceptibility on the relationship between ATTACHMENT SECURITY AND INTERNET GAMING DISORDER 35 attachment and IGD. It was initially predicted that insecure attachment and perceived stress would be significantly positively related to gaming addictions (Lemmens, Valkenburg & Gentile,
2015; Danet & Miljkovitch, 2017; Hill-Soderlund et al., 2008; Gander & Buchheim, 2015). No hypotheses were made pertaining to differences between attachment categories except that anxious and avoidant attachment groups were expected to exhibit higher gaming addiction scores and stress scores. It was also predicted that attachment would have a significant effect on gaming addiction scores through stress susceptibility.
The findings of this study extend those of the previous research pertaining to attachment and gaming addiction, and attachment and internet addiction (Danet & Miljkovitch, 2017;
Şenormancı et al., 2014; Eichenberg et al., 2017; Tavakoli, Jomehri & Farrokhi, 2014; Sroufe &
Waters, 1977; Hill-Soderlund et al., 2008; Gander & Buchheim, 2015; Gunnar et al., 1996;
Schore, 2003; Kaess et al., 2017). Results indicated that significant differences existed between the secure, dismissive-avoidant, fearful-avoidant and preoccupied groups in both gaming addiction scores and perceived stress. As expected, individuals with insecure attachment scored much higher on gaming addiction and perceived stress than those with secure attachment. This provides evidence for previous findings indicating that insecurely attached adults perceive more stress and are in-turn more susceptible to development of psychopathology (Hill-Soderlund et al.,
2008; Gander & Buchheim, 2015; Pietromonaco et al.’s, 2013; Schore, 2003). Additionally,
Individuals with preoccupied attachment styles reported the highest levels of trauma relative to the secure group. This was in line with evidence in the literature indicating that trauma may have negative effects on adaptive functioning that persist into adulthood (Raine et al.. 2001; Schore,
2003). findings of this study are also in concordance with findings by Muller, Sicoli & Lemieux
(2000) that individuals assessed for trauma (childhood abuse), who also reported posttraumatic ATTACHMENT SECURITY AND INTERNET GAMING DISORDER 36 stress symptoms, were predominantly found to have insecure attachment styles of attachment
(76%).
Additionally, when comparing the dismissive-avoidant, fearful-avoidant, preoccupied, and secure group, individuals who scored higher on the quadrant in anxiety (preoccupied, fearful-avoidant) showed the highest gaming addiction scores and perceived stress scores as measured by all the measures included in the methodology. Individuals who scored higher in avoidance showed lower scores on perceived stress and gaming addiction, indicating that attachment anxiety may play a role in worsening gaming addiction and stress outcomes. These findings were in agreeance with findings by Eichenberg et al. (2017) and Şenormancı et al.
(2014) that internet and gaming addiction tendencies significantly positively predicted self- reported anxious attachment scores.
Regression analyses further supported hypotheses, indicating that attachment type
(insecure & secure) significantly predicted gaming addiction above and beyond the covariates.
Additionally, it was found that attachment was also a significant predictor of perceived stress above and beyond the covariates. A significant causal relationship between perceived stress and gaming addiction was also supported by the regression analyses above and beyond the covariates. This supported findings by Kaess et al., (2017) that individuals with gaming addiction show greater levels of perceived stress and attenuated cortisol response. Perhaps, this may be indicative of a possible heightened vulnerability to stress in individuals with gaming addiction, in comparison to healthy controls.
Support for the mediation model was indicated by the hierarchical regression showing that attachment was a significant predictor of addiction while controlling for the perceived stress variable. Lastly, a PROCESS mediation analysis showed that perceived stress mediated the ATTACHMENT SECURITY AND INTERNET GAMING DISORDER 37 effect of attachment security on gaming addiction scores above and beyond the covariates, indicating that together, perceived stress and attachment accounted for a significant portion of the variance in gaming addiction scores. Findings suggest that development of insecure attachment of individuals may play a role in increasing vulnerability to stress, in-turn increasing one’s vulnerability to development of psychopathology. However, only partial mediation was supported by the model indicating that while perceived stress does explain a significant proportion of the effect of attachment on gaming addiction, it does not explain all of the effect as indicated by hierarchical regressions completed in the previous section. This indicates that further research is called for on other factors that may be related to development and etiology and gaming addiction.
Limitations and Future Directions
The findings of partial mediation may be due to the high significant correlation between attachment and perceived stress. According to Hoyle and Kenny (1999), full mediation indicates that there may be other models of mediation that may be consistent with the data. Data indicated that anxious attachment was significantly more correlated with gaming addiction (p = .01), than avoidant attachment (p = .01). Furthermore, anxious attachment was also more significantly correlated with perceived stress(p = .01), than avoidant attachment (p = .01) indicating that perhaps attachment anxiety may play a larger role in development of psychopathology than avoidant attachment, and that a moderation model should be assessed in the future.
Other limitations in this study include the significant differences in sample size per attachment category. The Dismissive-Avoidant group had a sample size of only 41, while each of the other attachment groups contained over 100 participants. Small sample size can often result in type-II errors when comparing across groups. Additionally, this disparity was also present ATTACHMENT SECURITY AND INTERNET GAMING DISORDER 38 when comparing the secure group (n = 109) to the insecure group (n = 314). The current study also had significantly more males (n = 295) than females (n = 127) than females, which could result in reduced generalizability of findings. However, findings that males scored higher than females on average in gaming addiction (p <.001), are consistent with the literature (Kaess et al.,
2017). Future studies should seek to assess differences between males and females in this mediation model and include equal sample sizes in the comparison groups.
Additionally, in this study only the secure and insecure groups were compared in the mediation analyses. Future studies should perhaps assess differences between all categories using a multi-categorical mediation model. Lastly, in this study, differences in reported trauma or adverse childhood experiences between attachment groups were assessed. However, adverse childhood experiences were not included in the mediation model. Exploratory analyses indicated that trauma explains a significant proportion of variance in the outcome variable, gaming addiction, and that it may play a role in moderating the relationship between attachment and gaming addiction, and attachment perceived stress. Therefore, future studies should also seek to assess the influence of trauma on the relationship between attachment and gaming addiction.
Conclusion
In conclusion, the findings of this study indicate that attachment security, as well as stress, may have significant implications in the understanding of the development and etiology of gaming addiction. Findings that both attachment and stress significantly predict gaming addiction, and that attachment may have a significant influence on gaming addiction through perceived stress provide support for the continued study of attachment security and stress as potential predictors and indicators of gaming addiction. Further research is called for on the ATTACHMENT SECURITY AND INTERNET GAMING DISORDER 39 pathways through which attachment may influence gaming addiction, and treatment of psychopathology.
ATTACHMENT SECURITY AND INTERNET GAMING DISORDER 40
References
Ainsworth, M. D. S. (1969). Object relations, dependency, and attachment: A
theoretical review of the infant-mother relationship. Child Development, 40(4), 969-1025.
doi:10.2307/1127008
American Psychiatric Association. (2013). Diagnostic and statistical manual of mental disorders
(5th ed.). Arlington, VA: American Psychiatric Publishing.
Amirkhan, J. H., Urizar, G. G., & Clark, S. (2015). Criterion validation of a stress measure: The
stress overload scale. Psychological Assessment, 27(3), 985–996.
doi:10.1037/pas0000081
Baumrind, Diana. (1966). Effects of authoritative parental control on child behavior. Child
Development, 37(4), 887-907. doi:10.2307/1126611.
Baron, R. M., & Kenny, D. A. (1986). The moderator–mediator variable distinction in social
psychological research: Conceptual, strategic, and statistical considerations. Journal of
Personality and Social Psychology, 51(6), 1173–1182. doi: 10.1037/0022-
3514.51.6.1173
Benarous, X., Morales, P., Mayer, H., Iancu, C., Edel, Y., & Cohen, D. (2019). Internet gaming
disorder in adolescents with psychiatric disorder: Two case reports using a developmental
framework. Frontiers in P sychiatry,10, 336. doi:10.3389/fpsyt.2019.00336
Bethell, C. D., Carle, A., Hudziak, J., Gombojav, N., Powers, K., Wade, R., & Braveman, P.
(2017). Methods to assess adverse childhood experiences of children and families:
Toward approaches to promote child well-being in policy and practice. Academic
Pediatrics, 17(7). doi: 10.1016/j.acap.2017.04.161
Bonnaire C, Phan O. (2017) Relationships between parental attitudes, family functioning and ATTACHMENT SECURITY AND INTERNET GAMING DISORDER 41
internet gaming disorder in adolescents attending school. Psychiatry Res., 255, 104- 110.
doi: 10.1016/j.psychres.2017.05.030
Bowlby, J. (1988). A secure base. New York: Basic Books.
Brand, M., Young, K. S., Laier, C., Wölfling, K., & Potenza, M. N. (2016). Integrating
psychological and neurobiological considerations regarding the development and
maintenance of specific Internet-use disorders: An Interaction of Person-Affect-
Cognition-Execution (I-PACE) model. Neuroscience & Biobehavioral Reviews,71, 252-
266. doi:10.1016/j.neubiorev.2016.08.033
Brennan, K. A., Clark, C. L., & Shaver, P. R. (1998). Self-report measurement of adult
attachment: An integrative overview. In J. A. Simpson & W. S. Rholes (Eds.),
Attachment theory and close relationships (p. 46–76). Guilford Press.
Bretherton, I. (1992). The origins of attachment theory: John Bowlby and Mary
Ainsworth. Developmental Psychology,28(5), 759-775. doi:10.1037/0012-1649.28.5.759
Cacioppo, M., Barni, D., Correale, C., Mangialavori, S., Danioni, F., Gori, A. (2019). Do
attachment styles and family functioning predict adolescents’ problematic internet use? A
relative weight analysis. Journal of Child & Family Studies, 28(5), 1263–1271.
doi:10.1007/s10826-019-01357-0
Caplan, S. E. (2003). Preference for online social interaction. Communication Research, 30(6),
625–648. doi: 10.1177/0093650203257842
Cohen, S., Kamarck, T., and Mermelstein, R. (1983). A global measure of perceived stress.
Journal of Health and Social Behavior, 24, 386-396. Retrieved from
http://www.jstor.org/stable/2136404
Cohen, H., Zohar, J., Gidron, Y., Matar, M. A., Belkind, D., Loewenthal, U., … Kaplan, Z. ATTACHMENT SECURITY AND INTERNET GAMING DISORDER 42
(2006). Blunted HPA Axis response to stress influences susceptibility to posttraumatic
stress response in rats. Biological Psychiatry, 59(12), 1208–1218. doi:
10.1016/j.biopsych.2005.12.003
Collins, N. L. (1996). Working models of attachment: Implications for explanation, emotion, and
behavior. Journal of Personality and Social Psychology, 71, 810–832. doi:10.1037/0022-
3514.71.4.810
Collins, N. L., & Read, S. J. (1990). Adult attachment, working models, and relationship quality
in dating couples. Journal of Personality and Social Psychology, 58(4), 644–663.
doi:10.1037/0022-3514.58.4.644
Davidson R. J., Fox N. A. (1989). Frontal brain asymmetry predicts infants’ response to maternal
separation. J. Abnorm. Psychol. 98(2), 127–131. doi:10.1037//0021-843x.98.2.127
Danet, M., & Miljkovitch, R. (2017). Monde virtuel : enjeux et risques liés à l’attachement.
Psychologie Francais, 62(1), 57-83. doi:10.1016/j.psfr.2015.10.002
Donovan W. L., Leavitt L. A. (1985). Physiologic assessment of mother-infant attachment. J.
Am. Acad. Child Psychiatry 24, 65–70. doi:10.1016/s0002-7138(09)60411-8
Eichenberg, C., Schott, M., Decker, O., & Sindelar, B. (2017). Attachment style and internet
addiction: An online survey. Journal of Medical Internet Research, 19(5).
doi:10.2196/jmir.6694
Elliott Christian, Martin Sellbom, and Ross B. Wilkinson (2019). Evaluating the Association
Between Psychopathy and Specific Attachment Models in Adults. Journal of Personality
Disorders, 33(1), 1-S16. doi:10.1521/pedi_2018_32_325
Felitti, V. J., Anda, R. F., Nordenberg, D., Williamson, D. F., Spitz, A. M., Edwards, V., … ATTACHMENT SECURITY AND INTERNET GAMING DISORDER 43
Marks, J. S. (1998). Relationship of childhood abuse and household dysfunction to many
of the leading causes of death in adults: The Adverse Childhood Experiences (ACE)
study. American Journal of Preventive Medicine, 14(4), 245-258. doi:10.1016/S0749-
3797(98)00017-8
Forsans, E. (2015, April 15). More than 150 million Americans play video games. Retrieved
from https://www.afjv.com/news/5038_more-than-150-million-americans-play-video-
games.htm
Fox N. A., Bell M. A., Jones N. A. (1992). Individual differences in response to stress and
cerebral asymmetry. Dev. Neuropsychol. 8, 161–184. doi:10.1080/87565649209540523
Fraley, R. C., Waller, N. G., & Brennan, K. A. (2000). An item response theory analysis of self-
report measures of adult attachment. Journal of Personality and Social Psychology,
78(2), 350-365. doi:10.1037/0022-3514.78.2.350
Gander, M., & Buchheim, A. (2015). Attachment classification, psychophysiology and frontal
EEG asymmetry across the lifespan: A review. Frontiers in human neuroscience, 9, 79.
doi:10.3389/fnhum.2015.00079
Gunnar, M. R., Brodersen, L., Nachmias, M., Buss, K., & Rigatuso, J. (1996). Stress reactivity
and attachment security. Developmental Psychobiology,29(3), 191-204.
doi:10.1002/(sici)1098-2302(199604)29:33.0.co;2-m
Griffiths, M. (2005) A ‘components’ model of addiction within a biopsychosocial framework,
Journal of Substance Use, 10(4), 191-197. doi:10.1080/14659890500114359
Griffiths MD, Pontes HM (2014) Internet addiction disorder and internet gaming disorder are
not the same. J Addict Res Ther, 5(4), e124, doi:10.4172/2155-6105.1000e124
Hayes, A. F., & Rockwood, N. J. (2017). Regression-based statistical mediation and moderation ATTACHMENT SECURITY AND INTERNET GAMING DISORDER 44
analysis in clinical research: Observations, recommendations, and implementation.
Behaviour research and therapy, 98, 39-57. doi: 10.1016/j.brat.2016.11.001.
Heimberg RG, Horner KJ, Juster HR, Safren SA, Brown EJ, Schneier FR, Liebowitz MR.
Psychometric properties of the Liebowitz Social Anxiety Scale. Psychological
Medicine, 29(1), 199–212. doi: 10.1017/s0033291798007879
Heinrich-Heine-Universität Düsseldorf (2019). 2G*Power: Statistical power analyses for
windows and mac. Retrieved from
http://www.psychologie.hhu.de/arbeitsgruppen/allgemeine-psychologie-und-
arbeitspsychologie/gpower.html.
Hill-Soderlund A. L., Mills-Koonce W. R., Propper C., Calkins S. D., Granger D. A., Moore G.
A., et al. . (2008). Parasympathetic and sympathetic responses to the strange situation in
infants and mothers from avoidant and securely attached dyads. Dev. Psychobiol. 50,
361–376. doi:10.1002/dev.20302
Kaess ,M., Parzer, P., Mehl, L,Weil, L., Strittmatter, E.,Koenig, J. (2017). Stress vulnerability in
male youth with internet gaming disorder. Psychoneuroendocrinology, 77, 244-251.
doi:10.1016/j.psyneuen.2017.01.008.
Kaufman, C. I., & Rosenblum, L. A. (1967). The reaction to separation in infant monkeys:
Anaclitic depression and conservation-withdrawal. Psychosomatic Medicine, 29(6), 648–
675. doi: 10.1097/00006842-196711000-00010
King D. L., Delfabbro P. H. (2016). The cognitive psychopathology of internet gaming disorder
in adolescence. J. Abnorm. Child Psychol. 44, 1635–1645.
doi:10.1007/s10802-016-0135-y
Konrath, S. H., Chopik, W. J., Hsing, C. K., & O’Brien, E. (2014). Changes in Adult Attachment ATTACHMENT SECURITY AND INTERNET GAMING DISORDER 45
Styles in American College Students Over Time: A Meta-Analysis. Personality and
Social Psychology Review, 18(4), 326–348. https://doi.org/10.1177/1088868314530516
Kurdek, L. A. (2002). On Being Insecure about the Assessment of Attachment Styles. Journal of
Social and Personal Relationships, 19(6), 811–834. doi:10.1177/0265407502196005
Liu, C.-Y., & Kuo, F.-Y. (2007). A study of internet addiction through the lens of the
interpersonal theory. Cyber Psychology & Behavior, 10(6), 799–804.
doi:10.1089/cpb.2007.9951
Lemmens, J. S., Valkenburg, P. M., & Gentile, D. A. (2015, January 5). The Internet Gaming
Disorder Scale. Psychological Assessment. Advance online publication, 27(2), 567-82.
doi:10.1037/pas0000062
Mann, L., Burnett, P., Radford, M., & Ford, S. (1997). The Melbourne decision making
questionnaire: an instrument for measuring patterns for coping with decisional conflict.
Journal of Behavioral Decision Making, 10(1), 1–19. doi: 10.1002/(sici)1099-
0771(199703)10:1<1::aid-bdm242>3.0.co;2-x
Monacis, L., Palo, V. D., Griffiths, M. D., & Sinatra, M. (2017). Exploring individual
differences in online addictions: The role of identity and attachment. International
Journal of Mental Health and Addiction, 15(4), 853-868. doi:10.1007/s11469-017-9768-
5
Nachmias, M., Gunnar, M., Mangelsdorf, S., Parritz, R. H., & Buss, K. (1996). Behavioral
inhibition and stress reactivity: The moderating role of attachment security. Child
Development,67(2), 508. doi:10.2307/1131829
Pallini, Susanna, et al. (2014). Early child–parent attachment and peer relations: A meta- ATTACHMENT SECURITY AND INTERNET GAMING DISORDER 46
analysis of recent research. Journal of Family Psychology, 28(1), 118–123.,
doi:10.1037/a0035736.
Pietromonaco, P. R., Debuse, C. J., & Powers, S. I. (2013). Does Attachment Get Under the
Skin? Adult romantic attachment and cortisol responses to stress. Current Directions in
Psychological Science,22(1), 63-68. doi:10.1177/0963721412463229
Pontes, H. M., & Griffiths, M. D. (2015). Measuring DSM-5 Internet Gaming Disorder:
Development and validation of a short psychometric scale. Computers in Human
Behavior, 45, 137-143. doi:10.1016/j.chb.2014.12.006
Raine, A., Park, S., Lencz, T., Bihrle, S., Lacasse, L., Widom, C., Louai, A-D. & Singh,
M. (2001), Reduced right hemisphere activation in severely abused violent offenders
during a working memory task: An fMRI study. Aggressive Behav., 27(2), 111–129.
doi:10.1002/ab.4
Rees C. (2007). Childhood attachment. The British journal of general practice : the journal of
the Royal College of General Practitioners, 57(544), 920–922.
https://doi.org/10.3399/096016407782317955
Şenormancı, O., Şenormancı, G., Güçlü O. (2014). Two cases of excessive internet use with
comorbid family relationship problems, Neurology and Neurosurgery, 51(3), 280-282.
doi: 10.4274/npa.y6939
Schore, A. N. (2003). Advances in neuropsychoanalysis, attachment theory, and trauma
research: Implications for self-psychology. In affect regulation & the repair of the self
(pp. 433–484). New York: W.W. Norton.
Schneider, L. A., King, D. L., and Delfabbro, P. H. (2018). Maladaptive coping styles in ATTACHMENT SECURITY AND INTERNET GAMING DISORDER 47
adolescents with Internet gaming disorder symptoms. Int. J. Mental Health Addict. 16,
905–916. doi: 10.1007/s11469-017-9756-9
Sibley, C. G., & Liu, J. H. (2004). Short-term temporal stability and factor structure of the
revised experiences in close relationships (ECR-R) measure of adult attachment.
Personality and Individual Differences, 36(4), 969-975. doi: 10.1016/S0191-
8869(03)00165-X
Siegel, D. J., & Hartzell, M. (2003). Parenting from the inside out: How a deeper
self-understanding can help you raise children who thrive. New York: Penguin.
Simpson, J. A. & Rholes W. S. (1998), Attachment theory and close relationships (pp. 46-76).
New York: Guilford Press.
Şenormancı, Ö, Şenormancı, G., Güçlü, O., & Konkan, R. (2014). Attachment and family
functioning in patients with Internet addiction. General Hospital Psychiatry, 36(2), 203-
207. doi:10.1016/j.genhosppsych.2013.10.012
Shin, S., Kim, N., & Jang, E. (2011). Comparison of problematic internet and alcohol use and
attachment styles among industrial workers in Korea. Cyberpsychology, Behavior and
Social Networking, 14(11), 665-72. doi: 10.1089/cyber.2010.0470
Smuts, A. B. (1985). History and research in child development: In celebration of the 50th
anniversary of the Society. Chicago, IL: Univ. of Chicago Pr.
Snyder, R., Shapiro, S., & Treleaven, D. (2012) Attachment Theory and Mindfulness. Journal of
Child and Family Studies, 21(5), 709–717. doi:10.1007/s10826-011-9522-8
Spielberger, C. D. (1989). State-trait anxiety inventory: Bibliography (2nd ed.). Palo Alto, CA:
Consulting Psychologists Press.
Spielberger, C. D., Gorsuch, R. L., Lushene, R., Vagg, P. R., & Jacobs, G. A. (1983). Manual for ATTACHMENT SECURITY AND INTERNET GAMING DISORDER 48
the State-Trait Anxiety Inventory. Palo Alto, CA: Consulting Psychologists Press.
Sroufe L. A., Waters E. (1977). Heart rate as a convergent measure in clinical and developmental
research. Merrill Palmer Q. 23(1), 3–27.
Sugaya, N., Shirasaka, T., Takahashi, K., & Kanda, H. (2019). Bio-psychosocial factors of
children and adolescents with internet gaming disorder: A systematic review.
BioPsychoSocial Medicine, 13(1). doi:10.1186/s13030-019-0144-5
Szepsenwol, O., Simpson, J. A., Griskevicius, V., & Raby, K. L. (2015). The effect of
unpredictable early childhood environments on parenting in adulthood. Journal of
Personality and Social Psychology, 109(6), pp. 1045–1067.
Tavakoli, N., Jomehri, F., & Farrokhi, N. A. (2014). The relationship between attachment styles
and social self-efficacy with internet addiction in Iranian Students. Bulletin of
Environment, Pharmacology and Life Sciences, 3(2), 302–307. doi: 2277-1808
Throuvala, M. A., Janikian, M., Griffiths, M. D., Rennoldson, M., & Kuss, D. J. (2019). The role
of family and personality traits in Internet gaming disorder: A mediation model
combining cognitive and attachment perspectives. Journal of Behavioral Addictions,1-15.
doi:10.1556/2006.8.2019.05
Waters, S. F., Virmani, E. A., Thompson, R. A., Meyer, S., Raikes, H. A., & Jochem, R. (2010).
Emotion Regulation and Attachment: Unpacking Two Constructs and Their Association.
Journal of psychopathology and behavioral assessment, 32(1), 37–47.
doi:10.1007/s10862-009-9163-z
Watson, D., Clark, L. A., & Tellegen, A. (1988). Development and validation of brief measures
of positive and negative affect: the PANAS scales. Journal of personality and social
psychology, 54(6), 1063. doi: 10.1037/0022-3514.54.6.1063 ATTACHMENT SECURITY AND INTERNET GAMING DISORDER 49
World Health Organization. (2018). International Classification of Diseases – 11.
Retrieved from https://icd.who.int/browse11/l-m/en#/http://id.who.int/icd/entity/
338347362
Wohl, M. J. A., Matheson, K., Young, M. M., & Anisman, H. (2008). Cortisol rise following
awakening among problem gamblers: dissociation from comorbid symptoms of
depression and impulsivity. Journal of Gambling Studies, 24(1), 79-90,
doi:10.1007/s10899-007-9080-6
ATTACHMENT SECURITY AND INTERNET GAMING DISORDER 50
Table 1 Descriptive statistics for gaming, stress, and demographic measures by attachment. ATTACHMENT CATEGORY
Secure Dismissive- Fearful-Avoidant Preoccupied Avoidant (n=109) (n=41) (n=129) (n=144) Variable Mean(SD) Mean(SD) Mean(SD) Mean(SD) Age (years) 30.72 (10.53) 32.37 (13.02) 28.93 (7.08) 31.15 (6.07)
Education (Likert) 5.15(1.29) 4.83(1.34) 5.82(1.39) 6.33(1.15)
Attachment 1.87(0.69) 2.50(0.78) 4.58(0.62) 5.32(0.59) Related Anxiety
Attachment 2.00(0.66) 4.47(0.58) 3.98(.435) 3.18(0.37) Related Anxiety
Gaming Addiction 6.13(4.47) 8.76(6.77) 13.23(6.33) 16.94(6.55)
Perceived Stress 51.21(23.90) 61.07(24.24) 86.19(24.07) 110.01(26.62)
Trauma 1.76(2.11) 2.51(2.83) 4.50(3.32) 6.76(2.94)
Note. Gaming addiction was measured using the IGDS (Lemmens, Valkenburge & Gentile, 2015); Attachment Related Anxiety was measured using the ECR-R (Fraley, Waller & Brennan, 2002); Perceived stress was measured using the SOS (Amirkhan, Urizar & Clark, 2015); Trauma was measured using the ACE (Felitti et al., 1998). Significant effect of Age, p’s < .066. Significant effect of Attachment Anxiety, p’s < .001 Significant effect of Attachment Avoidance, p’s < .001 Significant effect of perceived stress, p’s < .001 Significant effect of Gaming Addiction, p’s > .001 Significant effect of Trauma, p’s > .001
ATTACHMENT SECURITY AND INTERNET GAMING DISORDER 51
Table 2 Association of Gaming Addiction, Perceived Stress and Attachment IGDS IAT SOS ACE Attachment Attachment Last Score Score Anxiety Avoidance Gamed IGDS --- .678** .628** .690** .625** .256** .105* Score IAT --- .579** .550** .649** .354** .149** Score SOS --- .638** .756** .270** .196** Score ACE --- .607** .173** .193** Score Attachment --- .406** .214* Anxiety Attachment --- .089 Avoidance Last Gamed \ \--- Note. * .05 significance, ** .01 significance.
ATTACHMENT SECURITY AND INTERNET GAMING DISORDER 52
Table 3 Hierarchical regression Attachment and IGDS Step 1 Step 2 Step 1 ΔR2 FΔR2 df β p β p Sex .108 26.51 2, 419 5.28 <.001 4.07 <.001 Age .12 .004 .113 .002 Step 2 Attachment .298 60.53 3,418 -7.50 <.001
Note. Attachment was measured using the ECR-R (Fraley, Waller & Brennan, 2002).
ATTACHMENT SECURITY AND INTERNET GAMING DISORDER 53
Table 4 Hierarchical regression Attachment and Perceived Stress Step 1 Step 2 Step 1 ΔR2 FΔR2 df β p β p Sex .034 8.34 2, 419 14.69 <.001 8.08 .010 Age .043 .826 .028 .863 Step 2 Attachment .298 60.53 3,418 -41.08 <.001
Note. Attachment was measured using the ECR-R (Fraley, Waller & Brennan, 2002).
ATTACHMENT SECURITY AND INTERNET GAMING DISORDER 54
Table 5 Hierarchical regression Perceived Stress and IGDS Step 1 Step 2 Step 1 ΔR2 FΔR2 df β p β p Sex .108 26.51 2, 419 5.28 .001 3.41 <.001 Age .12 .004 .11 .001 Step 2 SOS_Total .439 110.65 3,418 .127 <.001
Note. Perceived stress was measured using the SOS (Amirkhan, Urizar & Clark, 2015).
ATTACHMENT SECURITY AND INTERNET GAMING DISORDER 55
Table 6 Hierarchical regression Perceived Stress, Attachment and IGDS Step 1 Step 2 Step 3 Step 1 ΔR2 FΔR2 df β p β p β p Sex .108 26.51 2, 5.28 .001 3.41 <.001 3.22 <.001 419 Age .12 .004 .11 .001 .110 <.001 Step 2 SOS_Total .439 110.65 3,418 .127 <.001 .105 <.001 Step 3 Attachment .462 91.40 4,417 -3.17 <.001 Note. Perceived stress was measured using the SOS (Amirkhan, Urizar & Clark, 2015); Attachment was measured using the ECR-R (Fraley, Waller & Brennan, 2002).
ATTACHMENT SECURITY AND INTERNET GAMING DISORDER 56
Table 7 Mediation Effects of Perceived Stress on the Relationship between Attachment and Gaming Addiction, N = 422 95% CI Effect b Lower Upper Total -7.50 -8.88 -6.12 Direct -3.17 -4.59 -1.74 Indirect (mediation) -4.33 -5.37 -3.34
Note. Perceived stress scores, and attachment together, significantly predicted gaming addiction scores when covariates were held constant. Attachment continued to significantly predict gaming addiction scores, even when perceived stress was included in the model indicating that full mediation did not occur.
ATTACHMENT SECURITY AND INTERNET GAMING DISORDER 57
Table 8 Model coefficients for Mediation effects of Perceived Stress on the Relationship between Attachment and Gaming Addiction, N = 422 Consequent Variables
M(Stress) Y(IGDS) Antecedents β SE p β SE p X (Attachment) a -41.08 3.23 <.001 c’ -1.32 0.72 <.001 M (Stress) ------b 0.11 0.01 <.001 C1(Sex) f1 8.084 3.11 0.010 g1 3.22 0.60 <.001 C2(Age) f2 0.028 0.16 0.863 g2 0.11 0.03 <.001 Constant iM 86.82 6.01 <.001 iY -1.32 1.40 0.347
R2 =0 .306 R2 = 0.467 F(3, 418) = 61.61 ,p < .001 F(4, 417) = 91.40, p < .001 Note. C= covariate, M = mediator, X = independent variable. The direct paths between attachment and perceived stress, perceived stress and IGDS, and attachment and IGDS were tested. The indirect path between attachment and gaming addiction, holding perceived stress as the mediator, was also tested. C1 and C2 = covariates held constant in the model.
ATTACHMENT SECURITY AND INTERNET GAMING DISORDER 58
Figure 1 Conceptual Diagram of Proposed Mediation Model
Note. C= covariate, M = mediator, X = independent variable, Y = Dependent variable; The direct paths between attachment and perceived stress, perceived stress and IGDS, and attachment and IGDS were tested; The indirect path between attachment and gaming addiction, holding perceived stress as the mediator, was also tested; C1 and C2 = covariates held constant in the model.
ATTACHMENT SECURITY AND INTERNET GAMING DISORDER 59
Figure 2 Statistical Diagram of Proposed Mediation Model
Note. C= covariate, M = mediator, X = independent variable, Y = Dependent variable; *p < .05, **p < .01, *** p < .001; With every unit increase in in attachment (0 = secure, 1 = insecure), there was a 41.08 unit decrease in perceived stress; With every unit increase in attachment, there was a 1.32 unit decrease in gaming addiction score through perceived stress; With every unit increase in perceived stress, there was a .11 unit increase in gaming addiction scores; The total effect of attachment on IGDS indicated that with every unit increase in attachment, there was a 7.50 unit decrease in gaming addiction.